Mail Not Routing, stuck in /var/spool/mqueue.in

[Julian Field]

I have just seen one of my servers hang, due to the Bayes database getting
corrupt. If I set "Debug = yes" and "Debug SpamAssassin = yes" then it
printed millions of error messages about a "partial write (number  of
665)". Presumably there was supposed to be a number before "of", which is
supposed to increment but doesn't. I switched off Bayes (use_bayes 0) to
work around the problem.

What's the chance there are some messages out there which are capable of
causing Bayes corruption due to them containing strange characters/strings?
Has anyone suffered this problem who is using SA 2.60rc3?

At 15:43 04/09/2003, you wrote:
>Not sure if this will help or not, but I have noticed MailScanner on my
>backup/test server hanging.  An strace of the process showed nothing.  But
>ls -l /proc/<pid>/fd/ showed that it had some of the SpamAssassin bayes
>database files opened.  This is running MailScanner version 4.22-5 and
>SpamAssassin version 2.54.
>
>The strange thing is that MailScanner had been hung for over a day!  I would
>have thought that MailScanner would have timed out SpamAssassin by then.
>Killing MailScanner and restarting would fix the problem for a while, then
>it would happen again.  I know that some people who are having this problem
>are not using SpamAssassin, but perhaps there is a problem in the time out
>code for SpamAssassin or Virus Checking?
>
>I'm not complaining.  My main MailScanner server is working, and I set
>"use_bayes 0" on my backup, and that seemed to clear its problem.  The
>backup is a 233 MHz with 128 Mb of ram, and has other functions besides
>scanning email, so it is taxed when MailScanner scans email.
>
>Anyways, maybe this helps - if not, just ignore.
>
>Jason
>
> > -----Original Message-----
> > From: Julian Field [mailto:mailscanner at ECS.SOTON.AC.UK]
> > Sent: Thursday, September 04, 2003 5:26 AM
> > To: MAILSCANNER at JISCMAIL.AC.UK
> > Subject: Re: [MAILSCANNER] Mail Not Routing, stuck in
> > /var/spool/mqueue.in
> >
> >
> > One of the things you could do to track it is this:
> >
> > In /usr/sbin/MailScanner, look for the "sub WorkForHours" and scatter
> >          print STDERR "Got to point 1\n";
> > statements through it (obviously changing the number). Then
> > set "Debug =
> > yes" and you should see this output. If it is pausing horribly at some
> > particular stage of processing a batch, then this should show
> > it up. Leave
> > all the spam checks disabled if you can, these take quite a
> > long time anyway.
> >
> > And check your /etc/sysconfig/i18n has no mention of "utf8"
> > in it. That's
> > important.
> >
> > At 10:26 04/09/2003, you wrote:
> > >Hi!
> > >
> > > > Without login access to somebody's machine which is
> > *reliably* suffering
> > > > from this problem, I'm a bit stuck.
> > >
> > >If it was happening _right now_ i would -love- to give a
> > login, but it
> > >isnt.
> > >
> > >One of the other people who is suffering from this wanting
> > to give Julian
> > >a login for looking into this ?
> > >
> > >Bye,
> > >Raymond.
> >
> > --
> > Julian Field
> > www.MailScanner.info
> > MailScanner thanks transtec Computers for their support
> >

--
Julian Field
www.MailScanner.info
MailScanner thanks transtec Computers for their support